Related Experiment Video
Updated: Jul 8, 2026
Analyzing Mitochondrial Morphology Through Simulation Supervised Learning
Published on: March 3, 2023
Deep learning of multiplexed mitochondrial morphology identified natural compound combinations against skin
Ruiqi Zheng1, Min Yu2, Jihua Wei3
1NMPA Center for Innovation and Research in Regulatory Science, School of Pharmacy, Zhejiang University, Hangzhou, 310058, China.
Background:
Skin photoaging is primarily driven by mitochondrial dysfunction. Although natural products have demonstrated protective effects against Ultraviolet (UV)-induced damage, efficient screening strategies for mitochondrial-targeting compounds remain limited. Artificial Intelligence (AI)-assisted high-content screening strategy offers a valuable approach to identify bioactive candidates and elucidate their mechanisms.
Purpose:
This study aimed to employe an AI-assisted high-content screening strategy to identify mitochondrial-enhancing natural compounds that prevent skin photoaging and to evaluate their protective efficacy.
Methods:
This study employed an AI-assisted high-content screening strategy to identify mitochondria-targeting compounds that mitigate photodamage. UVB-induced models were established in cells, zebrafish, and mice with UVB lamp irradiation. Mitochondrial morphology and function were quantified using JC-1, TMRM, and MitoTracker probes. Cell viability, SA-β-gal activity, and protein expression were assessed by CCK-8 assay, SA-β-gal staining, and Western blotting, respectively.
Results:
Using AI-assisted high-content screening, we identified quercetin, spermidine, adenosine, Vitamin K2, and Mirabilis jalapa extract as mitochondrial protective compounds. Two optimized combinations, CC-1 and CC-2, restored ATP production, TCA cycle flux, and mitochondrial morphology in UVB-exposed human dermal fibroblasts (HDFs), reduced SA-β-gal activity, and upregulated COLLAGEN I, LAMIN B1, and SIRT3 expression. Both combinations alleviated the UVB-induced senescence phenotype by upregulating SIRT3. In a UVB-exposed zebrafish tail amputation model, CC-2 markedly promoted fin regeneration. These findings were validated in UVB-irradiated mice, where CC-2 demonstrated superior efficacy.
Conclusion:
Using AI-assisted high-content screening, we formulated two optimized combinations that alleviate UVB-induced skin photoaging, with CC-2 exhibiting superior efficacy, highlighting its potential as cosmetic additive.
